Zaer Zaqy '28 - National Research and AsiaTechX Summit Presentation in Singapore

This past summer, I had the privilege of speaking at two major events: the AsiaTechX Summit in Singapore, where I shared perspectives on youth-driven AI education, and at MIT, where I delivered a keynote on the intersection of algorithms and society in the field of artificial intelligence. Both opportunities grew out of my work founding Singapore Youth AI (SYAI), a nonprofit dedicated to making AI literacy accessible. At SYAI, I lead a team of students and young professionals to design and deliver national-level programs. Over the summer, we launched our AI Aware campaign to spark conversations about responsible AI use, and we scaled up our bootcamps to reach more students across Singapore. Speaking on international stages while also running grassroots programs at home showed me how local voices can play a meaningful role in global conversations about technology.

Alongside this leadership work, I conducted research at Singapore’s national research laboratories, focusing on the intersection of Bayesian neural networks, causal inference, and probabilistic reasoning. My project investigated how learning systems operate under uncertainty, particularly the limits of inference when models are tasked with distinguishing correlation from causation. This required engaging with concepts from graphical models, information theory, and statistical learning, and exploring how trade-offs in complexity and representation affect model performance. The challenge lay not just in implementation, but in confronting open theoretical questions: how do we formalize reasoning about interventions, how do priors shape outcomes in high-dimensional settings, and what does it mean for a model to “understand” causality?

Through this work, I gained experience in experimental design, probabilistic programming, and theoretical analysis, but perhaps more importantly, I learned the discipline of navigating ambiguity in research. Much of the process involved debugging subtle inconsistencies, refining assumptions, and interrogating results to separate genuine signal from statistical noise. It was a summer that sharpened both my technical foundation and my appreciation for the deeper questions that drive machine learning research.